05 December 2007

So I get really upset by unsafe driving, and the inevitable consequence of unsafe driving, traffic accidents. If it were possible for you to drive dangerously but for everyone else to be safe, then that would be a personal choice I'd be willing to let you make. One might claim that good defensive driving will inure someone from the consequences of another driver's unsafe choices. Keep safe following distance, be aware of the traffic around you, and if someone does something dangerous, you have a safe way out. While this may be true (ignoring the truly reckless driver who simply collides with you on purpose), even if you are not involved in an accident, it can have negative effects on you. I have observed many relatively minor accidents causing huge congestion on LA's freeways.

But here's where I start to wonder. How huge is that congestion in real terms? How many person-hours does a fender bender in rush hour traffic actually cost? Would people drive more safely if they knew this?

I also wonder if people realize their driving habits are unsafe. I recently accepted a ride from a friend of mine who I had always considered thoughtful and to have good judgment. She drove like a maniac, leaving no following distance, changing lanes erratically and generally not paying attention to the traffic around her. Of course, nothing bad happened during the ride. We arrived safely and probably got to our destination quite a bit sooner than we would have if I were driving. Because consequences can be so severe for something going wrong on the road (e.g. death), it seems smarter to me to err on the side of caution, but maybe I'm being overcautious.

How severe of a problem is, for example, tailgating? I think it would be neat (and here comes the call for action) to 1) look into the questions about the cost to society of (minor) traffic accidents and 2) develop some kind of simulator for tailgating. The simulator idea could even be kind of fun. Let the user select a speed and following distance. Then display a view out the front of your car of the car in front of you. It could be scaled for distance. After a random, short interval the car in front would suddenly brake. The user would then have to click a button marked "brake." Based on the user's reaction, the simulator would calculate if a collision occurred and then display the personal cost of the accident followed by an estimate of the cost to society of the accident. Of course, if there was no accident the user wins this round and tries again.

A lot of research would need to get done in order to make something like that useful. Unfortunately, I have no idea where to start. Leave a comment if you know the answer to these questions, or if you have a hint where I should look.

4 comments:

Cosmonaut Zero said...

The simulator you mention reminds me of a system F1 drivers use to measure/work on their reaction speed to changes in the road (like a turn or a stopped car.) The system is a lot less fancy than your proposed one though. It consists of a row of lights that are set so as to be pointing away from the driver (the first light is right next to the driver, the second a bit further away and so on)The lights light up from the farthest to the nearest at some set speed to simulate something coming quickly at the user. It is the users job to press a button before the closest light to him is lit up. Hopefully that made some sense.

Ben said...

That's really pretty much exactly what I had in mind, but projected onto a 2d computer screen. This would, of course, deprive the user of binocular depth perception, but that provides surprisingly little extra information, so it would still be useful.

Unknown said...

I remember mythbusters had a thing on tailgating a semi truck to save on gas. Unfortunately LA is just a maze of freeways and cars. The odds of you driving near a crazy driver is higher because it is LA.

Cosmonaut Zero said...

http://starlab5.blogspot.com/ come visit my new and improved blog